IN THE HISTOLOGY LAB I ASKED DENIS TO PULL OUT CASES 25906-93 and 26704-94. I cleared the table to the right of the operating scope and placed my clipboard and pen. I took out two tubes of vinyl polysiloxane and positioned them, along with a small spatula, a tablet of coated papers, and a digital caliper accurate to.0001 inch.

Denis placed two cardboard boxes on the end of the table, one large and one small, each sealed and carefully labeled. I eased the lid from the larger box, selected portions of Isabelle Gagnon’s skeleton, and laid them out on the right half of the table.

Next I opened the smaller box. Though Chantale Trottier’s body had been returned to her family for burial, segments of bone had been retained as evidence, a standard procedure in homicide cases involving skeletal injury or mutilation.

I removed sixteen Ziploc bags and put them on the left side of the table. Each was marked as to body part and side. Right wrist. Left wrist. Right knee. Left knee. Cervical vertebrae. Thoracic and lumbar vertebrae. I emptied each bag and arranged the contents in anatomical order. The two segments of femur went next to their corresponding portions of tibia and fibula to form the knee joints. Each wrist was represented by six inches of radius and ulna. The ends of the bones sawed at autopsy were clearly notched. I would not confuse these cuts with those made by the killer.

I pulled the mixing pad toward me, opened one of the tubes, and squeezed a bright blue ribbon of dental impression material onto the top sheet. Next to it I squirted a white ribbon from the second tube. Selecting one of Trottier’s arm bones, I placed it in front of me and picked up the spatula. Working quickly, I mixed the blue catalyst and the white base, kneading and scraping the two squiggles into a homogenous goo. I scraped the compound into a plastic syringe, then squeezed it out like cake decoration, carefully covering the joint surface.

I laid the first bone down, cleaned the spatula and syringe, tore off the used sheet, and began the process anew with another bone. As each mold hardened I removed it, marked it as to case number, anatomical site, side, and date, and placed it next to the bone on which it had been formed. I repeated the procedure until a rubbery blue mold sat next to each of the bones in front of me. It took over two hours.

Next I turned to the microscope. I set the magnification and adjusted the fiber-optic light to angle across the viewing plate. Starting with Isabelle Gagnon’s right femur I began a meticulous examination of each of the small nicks and scratches I had just cast.

The cut marks seemed to be of two types. Each arm bone had a series of trench-like troughs lying parallel to its joint surfaces. The walls of the troughs were straight and dropped to meet their floors at ninety-degree angles. Most of the trench-like cuts were less than a quarter of an inch in length and averaged five hundredths of an inch across. The leg bones were circled by similar grooves.

Other marks were V-shaped, narrower, and lacked the squared-off walls and floors of the trench-like grooves. The V-shaped cuts lay parallel to the trenches on the ends of the long bones, but were unaccompanied in the hip sockets and on the vertebrae.

I diagrammed the position of each mark, and recorded its length, width, and, in the case of the trenches, depth. Next I observed each trench and its corresponding mold from above and in cross-section. The molds allowed me to see minute features not readily apparent when viewing the trenches directly. Tiny bumps, grooves, and scratches marking the walls and floors appeared as three-dimensional negatives. It was like viewing a relief map, the islands, terraces, and synclines of each trench replicated in bright blue plastic.

The limbs had been separated at the joints, leaving the long bones intact. With one exception. The bones of the lower arms had been severed just above the wrists. Turning to the bisected ends of the radius and ulna, I noted the presence and position of breakaway spurs, and analyzed the cross-sectional surface of each cut. When I’d finished with Gagnon, I repeated the whole process for Trottier.

At some point Denis asked if he could lock something up, and I agreed, paying no attention to his question. I didn’t notice the lab grow quiet.

“What are you still doing here?”

I almost dropped the vertebra I was removing from the microscope.

“Jesus Christ, Ryan! Don’t do that!”

“Don’t go bughouse, I just saw the light and thought I’d drop in to see if Denis was putting in overtime slicing up something entertaining.”

“What time is it?” I gathered the other cervical vertebrae and placed them in their bag.

Andrew Ryan looked at his watch. “Five-forty.” He watched me lift the bags into the smaller cardboard box and set the cover on top.

“Find anything useful?”

“Yup.”

I tapped the cover into place and picked up Isabelle Gagnon’s pelvic bones.

“Claudel doesn’t put much stock in this cut-mark business.”

It was precisely the wrong thing to say. I put the pelvic bones in the larger box.

“He thinks a saw’s a saw.”

I laid the two scapulae in the box and reached for the arm bones.

“What do you think?”

“Shit, I don’t know.”

“You are of the carpentry and grout gender. What do you know about saws?” I continued laying bones in the box.

“They cut things.”

“Good. What things?”

“Wood. Shrubbery. Metal.” He paused. “Bone.”

“How?”

“How?”

“How.”

He thought a minute. “With teeth. The teeth go back and forth and cut through the material.”

“What about radial saws?”

“Oh well, they go around.”

“Do they slice through the material or chisel through it?

“What do you mean?”

“Are the teeth sharp on the edge or flat? Do they cut the material or rip their way through it?”

“Oh.”

“And do they cut when they go back or when they go forth?”

“What do you mean?”

“You said the teeth go back and forth. Do they cut on the back or on the forth? On the push stroke or on the pull stroke?”

“Oh.”

“Are they designed to cut on the grain or across the grain?”

“Does that matter?”

“How far apart are the teeth? Are they evenly spaced? How many are on the blade? What’s their shape? How are they angled front to back? Is their edge pointed or squared off? How are they set relative to the plane of the blade? What kind of...”

“Okay, okay, I see. So, tell me about saws.”

As I spoke, I placed the last of Isabelle Gagnon’s bones in the box and tapped on the cover.

“There must be hundreds of different kinds of saws. Crosscut saws. Ripsaws. Pruning saws. Hacksaws. Keyhole saws. Kitchen and meat saws. Ryoba saws. Gigli and rod saws. Bone and metacarpal saws. And those are just the hand-powered ones. Some run on muscle, and some are powered by electricity or gas. Some move with a reciprocating action, some use continuous action, some move back and forth, some use a rotating blade. Saws are designed to cut different types of materials and to do different things as they cut. Even if we just stick to handsaws, which is what we’ve got here, they vary as to blade dimensions, and the size, spacing, and set of their teeth.”

I looked to see if he was still with me. He was, eyes as blue as the flame in a gas burner.

“What all this means is that saws leave characteristic marks in materials such as bone. The troughs they leave are of different widths and contain certain patterns in their walls and floors.”

“So if you’ve got a bone you can tell the specific saw that cut it?”

“No. But you can determine the most likely class of saw that made the cuts.”

He digested that. “How do you know this is a handsaw?”

“Power saws don’t depend on muscle, so they tend to leave more consistent cuts. The scratches in the cuts, the striae, are more evenly patterned. The direction of the cut is also more uniform; you don’t see a lot of directional changes like you do with a handsaw.” I thought for a minute. “Since there isn’t a lot of human energy required, people using power saws often leave a lot of false starts. And deeper false starts. Also, because the saw is heavier, or sometimes because the person working it is putting pressure on the object being cut, power saws tend to leave larger breakaway spurs when the bone finally gives.”

“What if a really strong person is working a handsaw?”

“Good point. Individual skill and strength can be factors. But power saws often leave scratches at the start of the cut, since the blade is already moving when it makes contact. Exit chipping is also more marked with a power saw.” I paused, but this time he waited me out. “The greater transfer of energy with power saws can also leave a sort of polish on the cut surface. Handsaws don’t usually do that.”

I took a breath. He waited to be sure I was actually through.

“What’s a false start?”

“When the blade first enters the bone it forms a trough, or kerf, with corners at the initial striking surface. As the saw moves deeper and deeper into the bone, the initial corners become walls and the kerf develops a distinct floor. Like a trench. If the blade jumps out, or is pulled out, before going all the way through the bone, the kerf that’s left is known as a false start. A false start contains all kinds of information. Its width is determined by the width of the saw blade and the set of its teeth. A false start will also have a characteristic shape in cross-section, and the teeth of the blade may leave marks on its walls.”

“What if the saw goes straight through the bone?”

“If the cut progresses all the way through the bone, the kerf floor can still partially be seen in a breakaway spur. That’s a spike that’s left at the edge of the bone where it finally breaks. Also, individual tooth marks may be left on the cut surface.”

I dug Gagnon’s radius back out, found a partial false start on the breakaway spur, and angled the fiber-optic beam across it.

“Here, look at this.”

He leaned over and squinted into the eyepiece, fiddling with the focus knob.

“Yeah. I see it.”

“Look at the kerf floor. What do you see?”

“It looks lumpy.”

“Right. Those lumps are bone islands. They mean that the teeth were set at alternating angles from the saw blade. That kind of set causes a phenomenon known as blade drift.”

He raised his head from the microscope and looked at me blankly. The eyepiece had left double rings, grooving his face like that of a swimmer with tight goggles.

“When the first tooth cuts into the bone it tries to align itself to the plane of the blade. It seeks the midline, and the blade goes along with that. When the next tooth enters the bone it tries to do the same thing, but it’s set in the opposite direction. The blade readjusts. This happens as each tooth comes along, so the forces acting on the blade change constantly. As a result, it sort of drifts back and forth in the kerf. The more set to the teeth, the more the blade is forced to drift. A very wide set causes successive teeth to drift so much that material is actually left in the midline of the kerf. Bone islands. Lumps.”

“So they tell you the teeth were angled.”

“Actually, they tell more than that. Since each directional change of a tooth is caused by the introduction of a new tooth, the distance between these directional changes can tell the distance between teeth. Since islands represent the widest points of bone drift, the distance from island to island is equal to the distance between two teeth. Let me show you something else.”

I withdrew the radius and inserted the ulna so that the cut surface at its wrist end was illuminated, then I stepped back from the microscope.

“Can you see those wavy lines on the cut surface?”

“Yeah. Looks kind of like a washboard, only curvy.”

“That’s called harmonics. Blade drift leaves those peaks and valleys on the wall of the cut just as it leaves bone islands on the floor. The peaks and islands correspond to the wide points in drift; the valleys and narrow aspects of the floor correspond to the points in drift when the blade is closest to the midline.”

“So you can measure these peaks and valleys like you do the islands?”

“Exactly.”

“How come I don’t see anything farther down in the kerf?”

“Drift occurs mostly at the beginning or end of a cut, when the blade is free, not embedded in bone.”

“Makes sense.” He looked up. The goggles were back.

“Can you tell anything about direction?”

“Of blade stroke or blade progress?”

“What’s the difference?”

“Direction of stroke has to do with whether the blade is cutting on the push or the pull. Most Western saws are designed to cut on the push. Some Japanese saws cut on the pull. Some can cut on both. Progress has to do with the direction the blade moves through the bone.”

“Can you determine that?”

“Yup.”

“So what do you have?” he asked, rubbing his eyes and trying to look at me at the same time.

I took my time answering, kneading the small of my back, then reaching for my clipboard. I flipped through my notes, selecting relevant points.

“Isabelle Gagnon’s bones have quite a few false starts. The kerfs measure about.05 inches in width and have floors that, in most cases, have some dip to them. Harmonics are present, and there are bone islands. Both are measurable.” I flipped a page. “There’s some exit chipping.”

He waited for me to go on. When I didn’t, he said, “What does all that mean?”

“I think we’re dealing with a handsaw with alternating set teeth, probably a TPI of 10.”

“TPI?”

“Teeth per inch. In other words, tooth distance is about a tenth of an inch. The teeth are chisel type, and the saw cuts on the push stroke.”

“I see.”

“The blade drift is extreme and there’s a lot of exit chipping, but the blade seems to cut efficiently by chiseling the material clear. I think it’s probably a saw designed like a very large hacksaw. The islands mean the set has to be pretty wide, to avoid binding.”

“That narrows it to what?”

I was pretty sure I knew what had made the cuts, but wasn’t ready to share my thoughts.

“There’s someone else I want to talk to before I reach a conclusion.”

“Anything else?”

I flipped to the first page of my notes, and summarized the observations I’d made.

“The false starts are on the anterior surfaces of the long bones. Where there are breakaway spurs, they’re on the posterior aspects. That means the body was probably lying on its back when it was cut up. The arms were detached at the shoulders, and the hands were cut off. The legs were removed at the hips and the knee joints were severed. The head was removed at the level of the fifth cervical vertebrae. The thorax was opened with a vertical slash that penetrated all the way to the vertebral column.”

He shook his head. “Guy was a real whiz with a saw.”

“It’s more complicated than that.”

“More complicated?”

“He also used a knife.”

I adjusted the ulna and refocused. “Take another look.”

He bent over the scope, and I couldn’t help noticing his nice, tight butt. Jesus, Brennan...

“You don’t have to press quite so hard against the eyepieces.”

His shoulders relaxed somewhat, and he shifted his weight.

“See the kerfs we’ve been talking about?”

“Uh-huh.”

“Now, look to the left. See the narrow slash?”

He was silent for a moment as he adjusted the focus. “Looks more like a wedge. Not square. It’s not as wide.”

“Right. That’s made by a knife.”

He stood up. Goggles.

“The knife marks have a definite pattern. A lot of them parallel the saw false starts, some even cross them. Also, they’re the only kind I see in the hip joint and on the vertebrae.”

“Meaning?”

“Some of the knife marks overlie saw marks and some are underneath, so the cutting probably came before and after the sawing. I think he cut the flesh with the knife, separated the joints with the saw, then finished with the knife, maybe disconnecting any muscles or tendons that still held the bones together. Except for the wrists, he went right into the joints. For some reason he just sawed the hands off above the wrists, going right through the lower arm bones.”

He nodded.

“He decapitated Isabelle Gagnon and opened her chest using just the knife. There are no saw marks on any of the vertebrae.”

We were both silent for a few moments thinking about that. I wanted all this to sink in before I dropped the bombshell.

“I also examined Trottier.”

The brilliant blue eyes met mine. His gaunt face looked tense, stretched, as he prepared it to receive what I was about to say.

“It’s identical.”

He swallowed and took a deep breath. Then he spoke very quietly. “This guy must run Freon through his veins.”

Ryan pushed off from the counter just as a janitor poked his head through the door. We both turned to look at him, and, seeing our somber expressions, the man left quickly. Ryan’s eyes reengaged mine. His jaw muscles flexed.

“Run this by Claudel. You’re getting there.”

“I’ve got a couple of other things I want to check out first. Then I’ll approach Capitaine Congenial.”

He departed without saying good-bye, and I finished repacking the bones. I left the boxes on the table and locked the lab behind me. As I passed through the main reception area, I noted the clock above the elevators: 6:30 P.M. Once again it was me and the cleaning crew. I knew it was too late to accomplish either of the last two things I’d planned to do, but decided to try anyway.

I walked past my own office and down the corridor to the last door on the right. A small plaque said, INFORMATIQUE, with the name LUCIE DUMONT printed neatly below.

It had been long in arriving, but the LML and LSJ were finally coming on-line. Complete computerization had been achieved in the fall of ’93, and data were continually being fed into the system. Current cases could now be tracked, with reports from all divisions coordinated into master files. Cases from years past were gradually being entered into the database. L’Expertise Judiciaire had roared into the computer age, and Lucie Dumont was leading the charge.

Her door was closed. I knocked, knowing there would be no answer. At 6:30 P.M. even Lucie Dumont was gone.

I trudged back to my office, pulled out my membership directory for the American Academy of Forensic Sciences, and found the name I was looking for. I glanced at my watch, quickly calculating. It would be only four-forty there. Or would it be five-forty? Was Oklahoma on Mountain or Central time?

“Oh hell,” I said, punching in the area code and number. A voice answered and I asked for Aaron Calvert. I was told, in a friendly, twangy way, that I was speaking with the night service, but that they’d be glad to take a message. I left my name and number and hung up, still not knowing into what time zone I’d been speaking.

This was not going well. I sat a moment, regretful that my burst of resolve hadn’t come earlier in the day. Then, undaunted, I reached for the receiver again. I dialed Gabby’s number and got no response. Apparently, even the answering machine had dropped out. I tried her office at the university, and listened to the line roll over after four rings. As I was about to hang up, the phone was answered. It was the departmental office. No, they hadn’t seen her. No, she hadn’t picked up her mail for a few days. No, that wasn’t unusual, it was summer. I thanked them and hung up.

“Strike three,” I said to the empty air. No Lucy. No Aaron. No Gabby. God, Gabby, where are you? I wouldn’t let myself think about it.

I tapped the blotter with a pen.

“High and outside.”

I tapped some more.

“Fourth and long,” I added, ignoring the mixed metaphor. Tap. Tap.

“D.Q.”

I leaned back and flipped the pen end on end into the air.

“Double fault.”

I caught the pen and sent it airborne again.

“Personal foul.”

Another launch.

“Time to switch to a new game plan.”

Catch. Launch.

“Time to dig in and hold the line.”

I caught the pen and held it. Dig in. I looked at the pen. Dig in. That’s it.

“Okay,” I said, pushing back my chair and reaching for my purse.

“Try batting from the other side of the plate.”

I slung the purse over my shoulder and turned out the light.

“In your face, Claudel!”